It's because the SQL servers can do funny things with regard to null fields and uniqueness. If I'm not mistaken, SQLite and PostgreSQL have different behavior: one considers to NULL rows to be the same, and one considers them to be different. The restriction in persistent is to prevent people from shooting themselves in the foot. I've wished the restriction wasn't there a few times, but the alternative (having strange SQL error messages) is even worse. When I've *really* needed this feature, I've created a newtype wrapper that represents Nothing as an empty string, that way I don't need to worry about SQL's funny definition of NULL. Michael On Sun, Oct 24, 2010 at 10:59 AM, Max Cantor <mxcantor@gmail.com> wrote:

Is it strictly necessary to have the restriction on nullable fields in the Persistent library?  Since Eq a => Eq (Maybe a) it shouldn't be strictly necessary right?

I haven't looked at the underlying code yet, just curious..

Max

_______________________________________________ web-devel mailing list web-devel@haskell.org http://www.haskell.org/mailman/listinfo/web-devel